gannon-storm-rtk-analysis¶
A reproducible analysis of how RTK-GNSS positioning error evolved across the Iowa / Illinois / Indiana / Ohio corridor during the May 10-12, 2024 Gannon G5 superstorm. The artifact is one of four in the HELIOS program (577 Industries' NASA SBIR Phase I, subtopic SPWX.1.S26A); it converts §1.3 of the submitted proposal — the Gannon anecdote — from a story into a citable result with code, data manifests, and figures any operator can examine.
Headline¶
Across 25 of 25 NGS CORS stations spanning the IA, IL, IN, OH corridor, the v1 climatological model produced 2D horizontal SPP error exceeding the agronomic 2.5 cm threshold for an aggregate of ~1,300 station-hours during May 10-12, 2024 (peak Kp=9.0, Dst minimum=-406 nT).
(See results/quantitative.md for the per-station breakdown on GitHub.)
What's in here¶
stations.py— 25-station catalog, truth coordinates from each station's day-131 RINEX header.fetch.py— polite NGS CORS RINEX downloader with on-disk caching.swpc.py— GFZ Kp + Kyoto Dst archive fetchers.positioning.py— v1 climatological 2D-error model. See methodology.analysis.py— orchestrator that ties fetch + indices + positioning + aggregates into one pipeline.plotting.py— four figures with data/method/timestamp footers.- Three Jupyter notebooks for interactive exploration (
notebooks/). - A 1,707-word blog post draft under
blog-post/. - Tests with >70% coverage; v1 ships at ~80%.
Reproducibility¶
bash
make all # fetch + analyze + plot, end-to-end
make fetch # cache RINEX + indices only
make analyze # aggregate without re-fetching
make test # pytest + coverage
Cold-cache full run: ~5-10 min on a typical broadband connection. Warm-cache run: <30 s.
Status¶
v0.1.0 — initial release. Pre-stable. RTK-style equipment transfer functions
and full PPP/RTK refinement are scheduled for v2 once the HELIOS fusion
engine (helios-fusion-engine) ships its companion training infrastructure.